Reed's views, was [e2e] Cannara's views

Alex Cannara cannara at
Sun Apr 22 09:32:21 PDT 2001

John, thanks again for a factual approach.  I've been away, dealing with
some family matters, so have no idea how many responses this has
provoked in e2e, but I did notice an Enet-address thread starting that
accurately explains the uniqueness built into all 802 MAC addressing
/DLC.  I was amused to see a response floating the idea that unique
addresses are somehow bad, so IP's misdesigned kludge is 'good'.  None
of my houses came with addresses on them, but all of their lots did! 
However, as with 802.5, it's perfectly possible for a driver to give any
Enet interface any address -- some Token-Ring shops used the Locally-
Administered Address option to simpy put the person's phone extension in
their address (great when looking for all those Token Ring problems with
a Sniffer(r).

It's interesting to porbe and see just how many bureaucratic defense
mechanisms get tickled enough to expose their illogic in the Internet
parish.  :]


John Day wrote:
> At 18:39 -0700 4/16/01, Cannara wrote:
> >John, thanks, I agree with you.  I was wrong to say "optimized" in reference
> >to 56k links, etc.  It was indeed an effort to look around and see what might
> >be done, given what was thought needed to do.  Xerox' attitude was indeed
> >unfortunate (as in so many things).  But interestingly enough, Noorda and his
> >main coder at Novell, adapted what Xerox would not share (or what Noorda
> >refused to be seen using) and at least solved the addressing problem early on.
> >So, in some instances, ideas were available to use that the Internet folks
> >avoided, whether because they had no interest in, or were biased against,
> >LANs, or because they were just focussed in a narrow problem space.
> I almost mentioned Novell.  In early 1974, there was a group in the
> ARPANet that was trying to do some of what became embodied in Novell.
> Katie Hafner's book says it was shut down because ARPA was afraid
> they were losing control of the net.  That was the end of all upper
> layer work.  Without any research money, no one was going to work on
> it.
> >
> >By the way, using SNA or the telcos as straw men is very misleading -- SNA was
> >an extension of IBM's mainframe-controller-slave religion and the telcos were,
> precisely.  IBM (and others) were characterized by big, expensive,
> master/slave, polling, half-duplex, synchronous, centralized, and
> full of special cases, etc.  The Net from the earliest prided itself
> on being small (at least smaller), cheaper, peer, event-driven, full
> duplex, asynchronous, decentralized, and elegant degenerate cases,
> etc.  We prided ourselves on being everything they weren't.  And as
> far as the phone companies go, they were much of the same list as the
> mainframe plus *connections*, and we certainly weren't that!!!  ;-)
> A friend who is writing a book on this stuff has a great quote from
> Metcalfe about how much it irritated him when some phone company guys
> made derogatory remarks about the Net at ICCC and how much it spurred
> him shove it in their face.
> Actually, it is very interesting how the mainframes (data comm) and
> the phone companies had developed complementary models that allowed
> them to co-exist and how the Net came along and not only declared war
> on both but put them all on a collision course.
> >and still are, interested in fitting bits into standard frames and charging
> SNA has no choice.  As I have said for years, you can always make a
> peer architecture hierarchical, you can't go the other way.  Had IBM
> built a peer architecture and subset it to be hierarchical for the
> 70's market, we would never have succeeded.  But they built
> themselves into a dead end.
> >for them -- remember ATM to the desktop, where they would own our wallplate
> >and the cable to it?  Fortunately desktop ATM collapsed of its own economic
> >weight and backbone ATM will as well.
> Remember it?  Remember X.25 where the equivalent of a TIP was PART of
> the network!! and where if you ever heard Louis Pouzin talk at the
> time, the European PTTs wanted to make it so the only computers you
> could attach to their networks were theirs, just like their phones.
> ATM was a walk in the park by comparison.
> But really, had the USING group continued and developed a wide area
> version of something like what Novell had. (And the thinking at the
> time was heading that way.)  Would we have put ourselves into a
> dead-end or would it have actually been a better place for the kind
> of mass adoption that we have seen?  In a sense, we have allowed the
> rest of the world to share the excitement we felt 30 years ago but
> are we really that much further along?
> Remember NLS on an IMLAC!  ;-)))  who needs Berners-Lee!  ;-))  to
> borrow from IM world: LOL.
> Take care,
> John

More information about the end2end-interest mailing list